Skip to main content
. 2022 Sep 12;17(9):e0274300. doi: 10.1371/journal.pone.0274300

Fig 3. Meme preprocessing.

Fig 3

A. Embeddings of image features and object tags in VinVl. B. Meme preprocessing by VinVl and OCR in parallel results in a triplet input of image feature, object tags, and text caption to OSCAR+.